A Novel Artificial Intelligence Algorithm of Synthetic Sampling for Boosting Accurate Prediction of Infrequent Health Outcomes.

8:20 AM–8:40 AM May 21, 2020 (Conference Time: US - Pacific)

8:20 AM–8:40 AM May 21, 2020

Regency B


Abstract: Real-world clinical health outcomes are usually not balanced or evenly distributed, (i.e. 20% of patients had stroke and 80% patients did not). Imbalanced (infrequent or rare) health outcomes have a significant negative impact on machine-learning model prediction performance. We developed a novel artificial intelligence algorithm that significantly improved a gradient boosting model’s prediction performance for rare health outcomes using a virtual patient cohort and real-world healthcare data.

Describe the new knowledge and additional skills the participant will gain after attending your presentation.: The attendees might be able to know a new AI algorithm that may significantly boost machine-learning prediction models’ performance for predicting infrequent or rare health outcomes.


Gang Fang (Presenter)
University of North Carolina

Izabela Annis, University of North Carolina